Search

The link between precipitation, drought indices, and NDVI is discussed in this paper. The data were processed on the example of the extreme eastern region of Georgia-Kakheti for the period 2016-2020. The study area has landscapes of natural (mixed and deciduous forests) and agriculture (vineyards, orchards, cereals, and vegetables). The NDVI was generated using Sentinel 2 images with a 10 m pixel resolution, and the average monthly NDVI was derived using Arc map 10.8. Drought indices (SPI and SPEI) were calculated according to the daily climate data from five rain gauges located in the study area in program R. Several trends emerged from the results. The correlation between NDVI, precipitation and drought indices vary according to natural and agricultural landscapes. A relatively low correlation was observed between the average monthly NDVI, precipitation and drought indices in the case of forests in Lagodekhi. These areas are relatively humid locations in Kakheti. In the southeast of the region, where arid forests are represented, NDVI was found to be more sensitive to precipitation and, consequently, drought indices. However, in contrast to the previously described locations, SPI and SPEI differed significantly from each other. In the case of agriculture landscapes, this connection is more complex and depends on the crop type and the vegetation period. [ABSTRACT FROM AUTHOR]